Anal. Chem. 1999, 71, 1378-1383
Preliminary Investigation of Electrothermal Vaporization Sample Introduction for Inductively Coupled Plasma Time-of-Flight Mass Spectrometry Patrick P. Mahoney, Steven J. Ray, Gangqiang Li,† and Gary M. Hieftje*
Department of Chemistry, Indiana University, Bloomington, Indiana 47405
The coupling of an electrothermal vaporization (ETV) apparatus to an inductively coupled plasma time-of-flight mass spectrometer (ICP-TOFMS) is described. The ability of the ICP-TOFMS to produce complete elemental mass spectra at high repetition rates is experimentally demonstrated. A signal-averaging data acquisition board is employed to rapidly record complete elemental spectra throughout the vaporization stage of the ETV temperature cycle; a solution containing 34 elements is analyzed. The reduction of both molecular and atomic isobaric interferences through the temperature program of the furnace is demonstrated. Isobaric overlaps among the isotopes of cadmium, tin, and indium are resolved by exploiting differences in the vaporization characteristics of the elements. Figures of merit for the system are defined with several different data acquisition schemes capable of operating at the high repetition rate of the TOF instrument. With the use of both ion counting and a boxcar averager, the dynamic range is shown to be linear over a range of at least 6 orders of magnitude. A pair of boxcar averagers are used to measure the isotope ratio for silver with a precision of 1.9% RSD, despite a cycle-to-cycle precision of 19% RSD. Detection limits of 10-80 fg are calculated for seven elements, based upon a 10-µL injection. Inductively coupled plasma mass spectrometry (ICPMS) has found wide acceptance as a powerful analytical method for elemental analysis in a range of applications. To further extend the capablities of ICPMS, or to overcome specific limitations of the technique, a number of different sample introduction techniques have come into routine use, including, among others, laser ablation, flow injection analysis, liquid and gas chromatographic separation, and electrothermal vaporization (ETV).1-4 In particular, the use of electrothermal vaporization for sample introduction presents several advantages over conventional solu† Present address: Hewlett Packard Laboratories, 3500 Deer Creek Rd., Building 26U, MS 26U-6, Palo Alto, CA 94303. (1) Montaser, A.; Minnich, M. G.; McLean, J. A.; Liu, H.; Caruso, J. A.; McLeod, W. In Inductively Coupled Plasma Mass Spectrometry; Montaser, A., Ed.; Wiley-VCH: New York, 1998; pp 83-264. (2) Ng, K. C.; Caruso, J. A. In Sample Introduction in Atomic Spectroscopy; Sneddon, J., Ed.; Elsevier: New York, 1990; Vol. 14, pp 165-193. (3) Gray, A. L.; Date, A. R. Analyst 1983, 108, 1033. (4) Carey, J. M.; Caruso, J. A. Crit. Rev. Anal. Chem. 1992, 23, 397-439.
1378 Analytical Chemistry, Vol. 71, No. 7, April 1, 1999
tion nebulization. Sample volume requirements are minimal (150 µL), higher sample transport efficiencies can be attained, and lower absolute and relative limits of detection can be achieved. Additionally, it has been shown that the appropriate ETV temperature program can be used to remove or separate the analyte from problematic matrix and solvent species which would otherwise cause spectroscopic and nonspectroscopic interferences.4-6 This ability to selectively volatilize species enables the analysis of samples with complex matrixes, in nonaqueous solvents, or with high dissolved salt or acid concentrations. Selective volatilization or atomization has also been employed as a means of speciation, separating different chemical forms of specific elemental species on the basis of differing thermal properties.7-9 The vaporization stage of the temperature program used with an ETV device generates a transient signal which typically has a duration of only a few seconds. When a scanned mass spectrometer is used to investigate such a transient, a compromise must be endured between the mass range to be examined and the accumulated signal for each element, as the total integration time available must be divided between the number of m/z values to be monitored.10 Thus, the limits of detection (LOD) and precision of the analysis are compromised when multiple elements must be determined. Because of the substantial cycle-to-cycle irreproducibility often encountered in the technique (2-10% RSD), the use of internal standard or isotope dilution techniques is often obligatory, techniques which may also be compromised in multielemental determinations. Further, different elemental species possess different volatilization characteristics and thus often “appear” at different times during the volatilization event.11-15 This differential volatilization may then create a spectral skew error, (5) Whittaker, P. G.; Lind, T.; Williams, J. G.; Gray, A. L. Analyst 1989, 114, 675. (6) Carey, J. M.; Evans, E. H.; Caruso, J. A.; Shen, W. L. Spectrochim. Acta, Part B 1991, 46, 1711-1721. (7) Robinson, J. W.; Weiss, S. Spectrosc. Lett. 1980, 13, 685-718. (8) Willie, S. N.; Gregoire, D. C.; Sturgeon, R. E. Analyst 1997, 122, 751-754. (9) Richner, P.; Wunderli, S. J. Anal. At. Spectrom. 1993, 8, 45-49. (10) Jarvis, K. E.; Gray, A. L.; Houk, R. S. Handbook of Inductively Coupled Plasma Mass Spectrometry; Chapman and Hall: New York, 1992. (11) Park, C. J.; Van Loon, J. C.; Arrowsmith, P.; French, J. B. Can. J. Spectrosc. 1987, 32, 29. (12) Richner, P.; Evans, D.; Wahrenberger, C.; Dietrich, V. Fresenius J. Anal. Chem. 1994, 349, 235-241. (13) Gregoire, D. C.; Lee, J. J. Anal. At. Spectrom. 1994, 9, 393. (14) Lamoureux, M. M.; Gregoire, D. C.; Chakrabarti, C. L.; Goltz, D. M. Anal. Chem. 1994, 66, 3217-3222. (15) Park, C. J.; Hall, G. E. M. J. Anal. At. Spectrom. 1987, 2, 473. 10.1021/ac9811625 CCC: $18.00
© 1999 American Chemical Society Published on Web 02/20/1999
an incorrect assesment of ion signal intensities as a result of the fact that a scanning mass spectrometer must investigate the different m/z values at different times during the time-dependent concentration profile of a transient signal.16 The result is, then, limited operation within a multielemental or scanning mode and the possibliity for erroneous quantitation.17 A mass spectrometer, or spectrophotometer,18-20 that provides simultaneous detection of all elements can be employed to eliminate this compromise between mass coverage and LOD/ precision for transient signals produced by ETV sample introduction. Furthermore, since the signals from all elements/isotopes are then collected at the same time, signal fluctuations and irreproducibility caused by the ion source, sample-loading procedure, or volatilization process can be eliminated through ratioing techniques and spectral skew limitations removed. With scanned systems, it has been shown that the effect of these fluctuations can be minimized and the precision improved by frequent peak hopping among several elements or isotopes.21,22 Hulmston and Hutton describe the use of a multichannel analyzer data aquisition system with a quadrupole ICPMS instrument to integrate an elemental mass spectrum during an ETV event, although 2 orders of magnitude in detectability and any temporal resolution is lost.23 To achieve real-time ratioing capabilities with a quadrupole instrument, Houk and co-workers have constructed an instrument consisting of two quadrupole mass filters in parallel.24-26 Double-sector instruments have similarly been used with multiple collectors to provide continuous, simultaneous monitoring of several elements or isotopes.27-30 However, these instruments have been designed for the observation of a limited number of m/z values, and the simultaneous monitoring of a large number of elements or isotopes in this manner may be impractical. Burgoyne et al.31 and Cromwell et al.32,33 have described doublesector instruments arranged in a Mattauch-Herzog geometry and equipped with array detectors to collect large sections of the elemental mass range simultaneously. These instruments have the potential to collect complete mass spectra in a continuous and simultaneous manner. Koppenaal and co-workers have described (16) Holland, J. F.; Enke, C. G.; Allison, J.; Stults, J. T.; Pinkston, J. D.; Newcome, B.; Watson, J. T. Anal. Chem. 1983, 55, 0003-2700. (17) Marshall, J.; Franks, J. At. Spectrom. 1990, 11, 177. (18) Tikkanen, M. W.; Niemczyk, T. M. Anal. Chem. 1985, 57, 2896-2900. (19) Tikkanen, M. W.; Niemczyk, T. M. Anal. Chem. 1984, 56, 1997-2000. (20) Tikkanen, M. W.; Niemczyk, T. M. Anal. Chem. 1986, 58, 366-370. (21) Vanhaecke, F.; Moens, L.; Dams, R.; Taylor, P. Anal. Chem. 1996, 68, 567569. (22) Furuta, N. J. J. Anal. At. Spectrom. 1991, 6, 199. (23) Hulmston, P.; Hutton, R. C. Spectroscopy 1991, 6, 35. (24) Allen, L. A.; Pang, H.; Warren, A.; Houk, R. S. J. Anal. At. Spectrom. 1995, 10, 267. (25) Allen, L. A.; Leach, J. J.; Pang, H. M.; Houk, R. S. J. Anal. At. Spectrom. 1997, 12, 171-176. (26) Warren, A. R.; Allen, L. A.; Pang, H. M.; Houk, R. S.; Janghorbani, M. Appl. Spectrosc. 1994, 48, 1360-1366. (27) Halliday, A. N.; Lee, D.; Christensen, J. N.; Walder, A. J.; Freedman, P. A.; Jones, C. E.; Hall, C. M.; Yi, W.; Teagle, D. Int. J. Mass Spectrom. Ion Processes 1995, 146/147, 21-33. (28) Walder, A. J.; Freedman, P. A. J. Anal. At. Spectrom. 1992, 7, 571-575. (29) Walder, A. J.; Abell, I. D.; Platzner, I.; Freedman, P. A. Spectrochim. Acta, Part B 1993, 48B, 397-402. (30) Walder, A. J.; Platzner, I.; Freedman, P. A. J. Anal. At. Spectrom. 1993, 8, 19-23. (31) Burgoyne, T. W.; Hieftje, G. M.; Hites, R. A. J. Am. Soc. Mass Spectrom. 1997, 8, 307-318. (32) Cromwell, E. F.; Arrowsmith, P. J. Anal. At. Spectrom. 1996, 7, 458-466. (33) Nam, S.; Montaser, A.; Cromwell, E. Appl. Spectrosc. 1997, 52, 161.
the use of a quadrupole ion trap for ICPMS.34-39 Although an ion trap is scanned during readout, no compromise between mass coverage and LOD or precision is necessary. All ions from the entire transient can be stored and subsequently analyzed, provided their number does not exceed the storage capacity of the ion trap. Because of the two-step trapping and read-out processes, however, the duty cycle of the ion trap may impose limitations when high temporal resolution of the transient signals is desired. Finally, Milgram et al.40 have recently coupled an ICP source to a Fourier transform ion cyclotron resonance mass spectrometer (FTICRMS). Like the ion trap, a FTICR-MS instrument is capable of storing all, or some subset of, ions from a transient signal and is also capable of detecting them in a truly simultaneous manner. In an analogous manner to the ion trap, however, the FTICR may also find limitations in ion storage capacity and duty cycle. The high repetition rate available with a time-of-flight (TOF) mass analyzer provides excellent temporal resolution and is exploited here for the temporal characterization of individual ETV transients. Myers et al. have described an ICPMS based upon a TOF mass spectrometer.41-46 Although a TOF mass analyzer does not detect ions of different m/z simultaneously, all ions within the extraction volume are injected into the drift tube at the same time. An entire mass spectrum is then recorded for each repetition cycle; because mass analysis is complete within a short period of time (50 µs), this cycle can be repeated at a high rate. Because all masses are sampled at the same instant, analysis can be conducted free from spectral skew and with the same precision, limits of detection, and resolution regardless of the number of m/z values to be investigated. The performance of an ICP-TOF instrument for transient signals has been reported previously by Mahoney et al. with the use of laser ablation.47 In the present study, this same TOF instrument is coupled to an ETV source. A high-speed analog-to-digital converter (ADC) board is employed to rapidly average and store successive TOF spectra, enabling multielement analysis to be performed with temporal resolution. Simultaneous multielemental determinations are accomplished by utilizing a set of compromise ETV operating parameters, facilitating the analysis of a solution containing 34 (34) Koppenaal, D. W.; Barinaga, C. J.; Smith, M. R. J. Anal. At. Spectrom. 1994, 9, 1053. (35) Eiden, G. C.; Barinaga, C. J.; Koppenaal, D. W. Rapid Commun. Mass Spectrom. 1997, 11, 37. (36) Eiden, G. C.; Barinaga, C. J.; Koppenaal, D. W. J. Anal. At. Spectrom. 1996, 7, 1161-1171. (37) Eiden, G. C.; Barinaga, C. J.; Koppenaal, D. W. J. Anal. At. Spectrom. 1996, 11, 317-322. (38) Barinaga, C. J.; Eiden, G. C.; Alexander, M. L.; Koppenaal, D. W. Fresenius J. Anal. Chem. 1996, 487-493. (39) Duckworth, D. C.; Barshick, C. M. Anal. Chem. 1998, 70, 709A. (40) Milgram, K. E.; White, F. M.; Goodner, K. L.; Watson, C. H.; Koppenaal, D. W.; Barinaga, C. J.; Smith, B. H.; Winefordner, J. D.; Marshall, A. G.; Houk, R. S.; Eyler, J. R. Anal. Chem. 1997, 69, 3714-3721. (41) Myers, D. P.; Mahoney, P. P.; Li, G.; Hieftje, G. M. J. Am. Soc. Mass Spectrom. 1995, 6, 920-927. (42) Myers, D. P.; Li, G.; Mahoney, P. P.; Hieftje, G. M. J. Am. Soc. Mass Spectrom. 1995, 6, 411-420. (43) Myers, D. P.; Li, G.; Mahoney, P. P.; Hieftje, G. M. J. Am. Soc. Mass Spectrom. 1995, 6, 400-410. (44) Myers, D. P.; Li, G.; Yang, P.; Hieftje, G. M. J. Am. Soc. Mass Spectrom. 1994, 5, 1008-1016. (45) Myers, D. P.; Hieftje, G. M. Microchem. J. 1993, 48, 259-277. (46) Mahoney, P. P.; Ray, S. J.; Hieftje, G. M. Appl. Spectrosc. 1997, 51, 16A28A. (47) Mahoney, P. P.; Li, G.; Hieftje, G. M. J. Anal. At. Spectrom. 1996, 11, 401.
Analytical Chemistry, Vol. 71, No. 7, April 1, 1999
1379
Table 1. Operating Conditions Plasma Parameters ICP forward power 1.1 kW ICP reflected power 1% RSD. Because each point on the plot represents an average of 300 measurements, better precision could be obtained with a greater number of averages, since thousands of spectra are available during the lifetime of each transient; for a 1-s transient, 16 600 spectra are available. To demonstrate quantitation capabilities, a simple isotopedilution experiment was performed. Two boxcar averagers were used to monitor the signals from the two isotopes of silver following the injection of a 500-pg aliquot of silver with and without a 56-pg spike of an isotopically enriched silver standard. The results are shown in Table 2. For each injection, a ratio was determined on the basis of the areas of the transients from the two isotope of silver. The ratios given in Table 2 are for the average of 10 such replicates; again, precision is limited by the data acquisition system. (55) Gregoire, D. C. Can. J. Anal. Sci. Spectrosc. 1997, 42, 1-9. (56) Shen, H.; Caruso, J. A.; Fricke, F. L.; Satzger, R. D. J. Anal. At. Spectrom. 1990, 5, 451.
The greatest sensitivity is obtained with the ion-counting data acquisition technique. Detection limits for seven elements are given in Table 3. Detection limits were calculated with the signal defined as the average peak area for five replicate injections of a 1-ppb solution of the element of interest. The noise was defined as the standard deviation of accumulated counts at the same m/z value for 10 replicate injections of a blank solution. The reported detection limits are limited by blank impurities and furnace memory effects and are comparable to those reported in the literature for similar analyses utilizing quadrupole-based ETVICPMS instruments.4,55,56 Although these values were determined in single-channel fashion, a multichannel counting system should permit determining them all in a single ETV cycle. CONCLUSION The combination of ETV and an ICP-TOF mass spectrometer has been explored. Although the performance of the instrument is currently limited by the data acquisition system, the abilities of ICP-TOFMS to perform complete elemental/isotopic analysis without sacrificing sensitivity, precision, or resolving power were demonstrated. A dynamic range of at least 6 orders of magnitude and detection limits of 10-80 fg were achieved; with a suitable data acquisition system, these values would be available for all elements and isotopes simultaneously. The temperature program of the furnace has been employed to alleviate both molecular and atomic isobaric overlaps on the basis of their different volatilities. ACKNOWLEDGMENT The authors thank the National Institutes of Health (Grant R01 GM 53560) for financial support and EG&G Instruments for the loan of the data acquisition equipment. P.P.M. thanks the Society of Analytical Chemists at Pittsburgh and Proctor and Gamble for fellowship support. Received for review October 21, 1998. Accepted January 5, 1999. AC9811625
Analytical Chemistry, Vol. 71, No. 7, April 1, 1999
1383